Characteristics of foreshocks and short-term deformation in the source area of major earthquakes

Abstract

We have examined the locations and radiation patterns of the foreshocks to the 4 February 1978 Haicheng earthquake. Using four stations, the foreshocks were located relative to a master event. They occurred very close together, no more than 6 kilometers apart. Nevertheless, there appear to have been too clusters of foreshock activity. The majority of events seem to have occurred in a cluster to the east of the master event along a NNE-SSW trend. Moreover, all eight foreshocks that we could locate and with a magnitude greater than 3.0 occurred in this group. The're also "appears to be a second cluster of foresfiocks located to the northwest of the first. Thus it seems possible that the majority of foreshocks did not occur on the rupture plane of the mainshock, which trends WNW, but on another plane nearly perpendicualr to the mainshock. plane. This inference is supported by differences among the radiation patterns of the foreshocks.We also examined the radiation patterns of the foreshocks, primarily through the amplitude ratios of P and S waves (Ap/As) recorded at each station. While the ratios were relatively constant at the more distant (A=100-200 kilometers) stations, two very different waveforms were recorded throughout the sequence at Yingkou (A=20 kilometers). Calculations of the amplitude ratios expected from the (not very different) fault plane solutions of the mainshock and some of the largest foreshocks showed that two different waveforms should be recorded at Yingkou. A definite correlation could not be made between the two mechanisms and the different location clusters. 1) IntroductionThe successful prediction of the 4 February, 1975 Haicheng earthquake (M=7.3) was due, in a large part, to the occurrence of a large f oreshock sequence of more than 500 events in the four days preceding the mainshock (Wu, 1976). Such a large sequence is a rare event -indeed, the Haicheng series is the only large foreshock sequence to have been well recorded -and we cannot hope that many other mainshocks will have such sequences. Many large earthquakes are preceded by a few foreshocks (Jones and Molnar, 1979) but so far we have been able to recognize them only in hindsigl}£. To increase the number of earthquakes that can be successfully predicted, we need to learn how to differentiate foreshocks from "normal" earthquake swarms before the mainshock occurs and when the shear number of events is not an overpowering consideration as it was at Haicheng. But precisely because ofits great length, the Haicheng sequence provides a unique opportunity to study this problem. The large number of events allows the characteristics of "this foreshock sequence to be seen more clearly. These characteristics could then be compared with those of earthquake swarms to the benefit of all people engaged in earthquake prediction research. Thus, the first part of the Chinese-American exchange in earth sciences has been a study of the Haicheng foreshock sequence resulting in this report.There are two purposes of this research. There is,...